CILFIT in Strasbourg

On 19 February 2024, the European Court of Human Rights decided not to answer the Estonian Supreme Court’s request for an advisory opinion on the basis of Protocol 16 (P16). For the first time, it dismissed a request because it did not concern a question of principle concerning the interpretation and application of ECHR rights. In doing so, the Grand Chamber made clear that it sees no role for itself in relation to easy questions that merely require a reiteration of its “well-established case law”. The decision is significant because the ECtHR provides clear contours as to what types of questions courts should (not) ask. The clair and balanced approach is to be welcomed. This post will explain why and reflect more generally on the functioning of P16, which entered into force 5.5 years ago and has led to nine requests to date.

What (not) to request?

The admissibility criteria are reminiscent of the conditions used by the Court of Justice of the European Union (CJEU) in relation to the preliminary ruling procedure in Article 267 TFEU. Article 1 of P16 stipulates that the request should be made 1) by a mandated highest court or tribunal; 2) in relation to a question of principle about the interpretation of the application of the ECHR; 3) in a case pending before the requesting national court and; 4) with reasons for the request and providing the relevant legal and factual background of the pending case. A panel of the Grand Chamber consisting of five judges decides whether the request should be accepted, after which the Grand Chamber decides on the merits. In contrast with Article 267 TFEU, a request under P16 is always optional and limited to the highest courts and tribunals only.

The Estonian request is the first time that the ECtHR explicitly dismissed an entire request on the basis of the second criterion. The Court reiterated its previous pronouncement that “questions of principle” warrant guidance from the Court given “their nature, degree of novelty and/or complexity”. It subsequently provided quite an extensive overview of its case-law in relation to the specific question of the respective case concerning ne bis in idem and concluded that the question is the subject of “well-established case-law of the Court”. This reasoning reflects the acte clair doctrine accepted in CILFIT, albeit the CJEU constructs this doctrine as an exception to the obligation of the highest courts to refer a preliminary question, not as an admissibility criterion.

The CJEU disposes of straightforward references on the basis of a reasoned order, following Article 99 of the Rules of Procedure. This provision essentially codifies the acte clair and (acte éclairé) doctrine. Hence, the CJEU does not declare straightforward references inadmissible, but renders a decision that at first sight does not differ considerably from its “normal” judgments. Yet, it does so much faster, without an Opinion of an Advocate General (AG) and in a three-judge formation (see, for example, here). By contrast, the ECtHR uses the “well-established case-law” as an inadmissibility ground. In practice, however, the approach is not very different from the reasoned orders used by the CJEU because the ECtHR’s decision contains nine substantive paragraphs with quite extensive considerations. Accordingly, the ECtHR still provided the Estonian court with a useful overview while it also set clear limits as to the invocation of P16. Therefore, it sidestepped the need to reprimand the Estonian court for asking a question it should have been capable of answering independently.

In September 2023, the ECtHR updated its Guidelines and provided additional useful guidance for the highest national courts when to issue a request for an advisory opinion. It noted that a request is primarily warranted when it involves “a novel point of Convention law”, when “the facts of the case do not seem to lend themselves to a straightforward application of the Court’s case-law” or in a case where it “appears to be an inconsistency in the case-law”. Interestingly, the guidelines provide the request of the Finnish Supreme Court as an example of a best practice because the Finnish court clearly set out the ambiguities and missing standpoints of the respective case-law.

The ECtHR did not turn down a request for the first time. It has previously rejected a request from the Slovakian Supreme Court in relation to effective criminal investigations based on the third condition, namely, the absence of a link with the pending case, as required under Article 1(2) P16. In the said case, the Slovakian Supreme Court referred mainly to Articles 2 and 3 ECHR, while the pending case dealt with Article 6 ECHR. In two other requests, the ECtHR only partly answered the questions posed. The Grand Chamber decided not to answer two of the four questions posed in an Armenian case about non-retroactivity for the same reason as in the Slovakian case, i.e. no link with the pending case. The Grand Chamber also refrained from engaging with the first question in the Lithuanian case about the impeachment of an MP because this question related to the execution of (prior) Strasbourg judgments. Henceforth, the ECtHR implicitly suggested that the question lacked the principled character required under Article 1(1) P16.

Between application and interpretation

One must welcome the clarity regarding the third admissibility requirement for two following reasons. First, P16-requests appropriate the limited resources of the Court facing significant workload challenges. Requests receive priority and keep seventeen judges of the Grand Chamber busy, inevitably at the expense of other cases. The Court should be commended for the efficient handling of the requests. The two “inadmissible” requests were rejected within less than three months, while four requests were decided between five and ten months. Three other requests were decided between fifteen and seventeen months. This is still below the average time of 17.3 months (2022) in Luxembourg. In this light, it is a blessing in disguise that the ECtHR has only received nine requests from courts in seven states, especially because it is far from certain that P16 would help in reducing the ECtHR’s workload, as a 2012 position paper predicted in a wishful-thinking manner. Courts in the other fifteen states that have ratified P16 have not made a request so far. One might doubt whether the number of requests will increase all too rapidly in the near future because of the preference (and legal obligation!) of national courts for a reference to the CJEU (see Opinion 2/13, paras. 198-199 and Article 5 of the Draft accession agreement). The focus on Luxembourg might be reinforced by the (recent) expansion of the EU legislation affecting or involving fundamental rights (e.g. the GDPR, the AI Act and harmonization in the criminal area).

The second reason concerns the relatively easy, arguably clair, questions in the majority of the P16-requests. The requesting courts seemed primarily interested in obtaining external and authoritative support in politically sensitive matters (a so-called “sword”, as I argued with Lize Glas before). One example is the (first) Armenian request in relation to the prosecution of the former President. Another is the Lithuanian request in relation to the impeachment of an MP in which the Court had previously determined a violation. In another Armenian case, the Court subtly noted that it is primarily for national courts “to resolve problems of interpretation of domestic legislation”.

These three cases illustrate that some courts have primarily used P16 requests to obtain case-specific guidance (or support) instead of an (abstract) interpretation of a legally complex subject matter. This is as such not surprising. Empirical research on the preliminary reference procedure under Article 267 TFEU and P16 shows that national courts appreciate concrete guidance and – in the words of Tridimas – outcome-oriented cases. The setup of P16 allows (or perhaps even encourages) case-specific questions. While P16 requires a question of principle, it also makes clear that the question should not be abstract, broad, or general but attached to a concrete pending case. Hence, this is an inherent tension in the setup of P16. The ECtHR needs to resist the temptation of dancing to the pipes of national courts. So far, it has done so in quite an elegant way by continuously emphasising the division of tasks and making clear that it does not have jurisdiction to assess the facts of the case or the merits of the positions of the parties given the division of tasks in P16 (e.g. para. 18 in P16-2020-001). The success of the ECtHR, by and large, rests on its criticism of being too abstract. Moonen and Lavrysen, for instance, held that the answers in relation to the first Armenian request were “not particularly useful” because national courts “should figure out the Convention issues for themselves”.

The aforementioned tension reflects a discussion in EU law about the interpretation versus application of EU law in the context of the preliminary ruling. Even though the CJEU employs a division of tasks similar to the ECtHR, it has at times stepped into the domain of the national judges by not limiting itself to only interpreting EU law but also applying the interpretation to the national legal or factual context in case-specific judgments. Several AGs criticized the CJEU for its willingness to go along with such wishes for “factual jurisprudence” (Bobek), while also calling for “self-restraint” from national courts (Jacobs in Wiener). The CJEU has partly responded to this criticism by limiting the obligation to submit a request for a preliminary reference in Consorzio Italian Management (often referred to as CILFIT 2.0) to questions about the correct interpretation of EU law. In that way, the CJEU affirms its final authority to define the correct interpretation of EU law but not necessarily the correct domestic application.

Walking the fine line between being strict and rude

While Strasbourg has set clear limits in its latest decision in relation to the Estonian request, some might argue that its approach is still too welcoming. It remains to be seen whether the Grand Chamber will introduce further admissibility restrictions should it receive more requests in the future, as predicted more than a decade ago. The practice of its Luxembourg counterpart indicates that it might become more restrictive over time as it receives more requests. The ECtHR needs to walk the fine line between being strict and rude. Especially considering the subsidiary nature of the ECHR system, it should not upset its cooperative relations with (most) of the national courts by dismissing their requests all too easily. The inadmissibility decision in the Estonian case, with an elaborate and useful discussion of the “well-established case-law” maintains this careful balance.